Microsatellite instability (MSI) is an important measure of genome instability in tissues (usually tumors). It has been related to cancer susceptibility and progression and is attributed to defects in mismatch repair (MMR) genes in hereditary non-polyposis colon cancer (HNPCC). MSI is generally determined by conducting PCR across a microsatellite locus and observing fragments (or alleles) that have different numbers of repeats than the progenitor alleles -- i.e., mutant alleles. In order to be detected by standard genomic PCR, any one mutant fragment must be present at substantial frequency (greater than 30 percent) -- a remarkable mutant frequency by any somatic cell genetic standard. It is our hypothesis that there are of a wide variety of perturbations in genes associated with repair and DNA metabolism, other than those resulting in complete loss of function of MMR genes. Such events could result in substantial levels of MSI frequencies (less than 0.3 - greater than 0.05), not observable by standard PCR, but which predispose to cancer. Therefore, we contend, the presence of such an MSI phenotype has considerable clinical significance but goes undetected by present methodology. A sensitive and efficient method for detecting and quantifying the MSI phenotype down to levels of approximately 0.05 mutant frequency at specific microsatellite loci has been developed -- small poo1 PCR coupled with multiplex GENESCAN analysis. Here we will apply SP/PCR to the two classes (MSI-high vs. MSI-low or stable) of HNPCC tumors and constitutive tissues as well as a limited set of "sporadic" colorectal patient materials and compare those results to age, allele size, and gender matched controls. We expect to determine and stratify the MSI phenotype in tumor material relative to MMR genotype; determine if there is an MSI phenotype in constitutive tissues that would be useful in presymptomatic identification of tissues and individuals at risk for developing cancer; and determine the range of variation of the MSI phenotype in the general population relative to size of progenitor allele, gender, and age.